An Interview with Philip Steadman

Curiously enough, one of the most widely read and heatedly discussed books about Vermeer in recent years, Vermeer's Camera: Uncovering the Truth behind the Masterpieces was not written by art historian, but by an architect, Philip Steadman. The controversial subject of this book is Vermeer's use of the camera obscura - a kind of predecessor to the modern photographic camera - as an aid to his painting. Since American photographer Joseph Pennell in 1891 first supposed that Vermeer employed the camera obscura the subject has been widely debated among art historians. However, not until Steadman's book has the topic been so thoroughly researched. Steadman painstakingly develops his argument through careful study of the history of the camera obscura, an exploration of 17th-century optics, and a detailed study of the light, optics, perspective, and measurement of a series of Vermeer's paintings. He goes to remarkable lengths to reconstruct Vermeer's studio and its furnishings, down to the angle of the light from its windows. The science is complex, but always clearly explained. Steadman's arguments are so carefully articulated and rational tested, that he has probably come as close as theoretically possible to proving that the camera obscura was in fact an integral part of the elusive Dutch master’s working method.

April 25, 2003

The Essential Vermeer: How did you arrive at the study of Vermeer and the camera obscura?

Prof Steadman: I was teaching perspective drawing to a group of summer school students in the 1970s, and decided to set them an exercise in reconstructing the three-dimensional scenes shown in paintings, using a method of ‘reverse perspective’. We chose 17th century Dutch interiors, including some by Vermeer, because the rooms have tiled floors. From there I became interested in the question of whether Vermeer shows the ‘same’ room or rooms repeatedly in his pictures. Only later did I come to realize that this question might have a bearing on his use of the camera obscura. So my route towards the research reported in Vermeer’s Camera was a circuitous one. Just recently I remembered, for the first time in many years, that we had a reproduction of ‘Girl with a Pearl Earring’ at home when I was a boy; so perhaps she provided some kind of long-term unconscious inspiration.

Have you significantly revised any of the ideas contained in your book ?

I stand by all the main arguments of the book. There are a few small mistakes that I would correct in a new edition. More important is the fact that the book has provoked a wonderful correspondence with historians, painters, film-makers, computer graphics specialists and Vermeer lovers from all over the world, who have offered criticisms, ideas and suggestions leading in many new directions. I have tried to describe some of these in an essay ‘Vermeer’s Camera: afterthoughts, and a reply to critics’ published on my web-site at

There were two significant points on which I remained dubious and undecided in the book, and offered different possibilities. The first was whether the room with the ‘squares and circles’ patterns of leading in the windows - which appears many times in pictures painted from the late 1650s onwards - is or is not the first floor studio used by Vermeer in Maria Thins’s house on the Oude Langendijk. In the book I thought there was insufficient evidence to come to a conclusion about this. But I was wrong.

In 2002 Ab Warffemius published his very convincing reconstruction of Maria Thins’s house, based on the 1675-8 pictorial map of Delft, a 19th century cadastral map giving the dimensions of the site, a drawing by the 18th century topographical artist Abraham Rademaker, and the probate inventory of the house and its contents made after Vermeer’s death. (See ‘Jan Vermeers huis: Een poging tot reconstructie’, Delfia Batavorum: Historisch Jaarboek voor Delft 2001, pp.60-78). Warffemius obtains dimensions for Vermeer’s studio that are extremely close to those which I derived – quite independently and on the basis of completely different sources, i.e. the paintings themselves – from my perspective calculations. Warffemius also shows windows and ceiling beams as in my reconstructions. (His arguments here have to do with the building’s structure, and the evidence of the Rademaker view, and not just with what is shown by Vermeer.) I think all this leaves little doubt that the paintings were made in, and depict, this actual room. A second reconstruction of the house by Henk J Zantkuijl, published more recently on Kees Kaldenbach’s web-site at seems to me to be less convincing. I have explained my reasons for preferring Warffemius’s proposals, in an exchange of views with Kaldenbach, again published on his web-site.

A second point on which I was undecided was the method by which Vermeer might have obtained images that were not mirrored right-to-left, relative to the room we see in the paintings. I inclined towards the idea that he employed a simple form of cubicle-type camera, in which the projected image would indeed have been mirrored (and upside-down), as shown in my Figure 52 (p.105); but that he then reversed this image again in the process of transferring it to canvas. I rejected – with some regret – the possibility that Vermeer’s actual room was not as we see it in the pictures, but was itself mirrored, like the room Alice steps into through the looking-glass in Lewis Carroll’s story (p.111).

More recently however I have come to appreciate that the very same room could present the appearance of a ‘looking-glass room’, by the simple expedient of viewing it from the opposite end. The windows would then be on the right of the view, instead of on the left as they are in the paintings. (The appearance of the tiled floor and the beamed ceiling would be unchanged, since they are symmetrical.) Vermeer could have placed his camera cubicle at this other end of the room. The arguments for and against this idea are set out in more detail in my web-site essay. The crucial advantage of this position and arrangement for Vermeer’s camera is that the projected images of the room on the camera screen are then exactly as in the paintings (although still upside-down). The room as a whole is ‘mirrored’; but then the camera lens reverses the image back again. Thus Vermeer could have traced and perhaps even painted over such images directly onto his canvas.

One of the crucial and most  heatedly debated points of your book is that Vermeer not only used the camera obscura as a means to observe reality from a more   "optical" perspective and as an aid to composition, but actually projected and subsequently traced the camera's image on to canvas. Do you feel that the evidence you have analyzed points concretely to further physical application of the camera by Vermeer after this point?

The central finding of my book is that the ‘projected images’, at the back wall, of at least six of Vermeer’s interiors, are the same sizes as the respective canvases. It is very hard to see how this strange geometrical phenomenon might have arisen, other than through the painter tracing those images directly in a camera. It is this discovery that has convinced a number of art historians, including some who previously were strongly opposed to any camera theories in relation to Vermeer. The geometrical nature of this argument gives it, I would suggest, a special force. I argue on this basis, then, that Vermeer used the camera to study optical images and effects; as an instrument with which to set up and adjust his compositional arrangements of sitters and furniture; and as a means to obtain traced outlines for complete pictures, at their final sizes.

You ask whether his use of the apparatus might have gone beyond this point: by which I take it that you are asking, ‘Might he actually have painted inside the camera?’ Several writers on Vermeer and optics have hinted at this, or even proposed it explicitly, and I am guilty here myself. But there is a serious problem, especially in relation to booth-type cameras in which one is enclosed in near-darkness; which is that one cannot then see the colours of one’s pigments.

On the other hand, it may be that Vermeer was able to paint in monochrome directly over the projected image in the camera. Vermeer’s first paint layers are in monochrome, either dark brown or blue over a light ground. Among the reasons why scholars have supposed Vermeer to be a camera user is his ‘uncannily true sense of tone’ (Kenneth Clark's phrase). There are methods by which the brightness of luminous patches within the projected image can be matched in grisaille. Perhaps here is the secret of Vermeer’s tonal precision?

Do you have any plans for further research in the field of the camera obscura or Vermeer's painting? If so, would you be so kind as to indicate the direction of your study?

I have in fact been carrying out some more research over the last year. I have built a large cubicle-type camera obscura in a room in my house in France. This room is roughly the same length and height as Vermeer’s studio, and has a similar beamed ceiling – although it has only one window, where Vermeer’s room has three. I have been experimenting with tracing, and painting in monochrome. The lens in the camera is 10 centimetres in diameter, in order to admit enough light to produce a bright image. The result is that depth of field is limited, and it is necessary to re-focus at different depths in the scene. I believe it is possible to see the consequences of these problems, in Vermeer’s compositions and painting technique.

Professor Martin Kemp of Oxford University and I are hoping to do some more extensive research on optical methods in 17th-century painting. In relation to Vermeer, I would be looking at the painter’s compositional procedures as these might have been affected by a camera technique; I would be making more practical experiments with tracing and painting in the camera; and I would be investigating Vermeer’s lighting effects, through studies with scale models and computer simulations. Should this work go ahead, Kemp and I would be very happy to keep the Essential Vermeer site in touch with developments.

Do you feel that Vermeer's use of the camera obscura as described in your book necessitates a revision of the artist's  expressive intentions or of his place within Dutch seventeenth-century philosophic, scientific or cultural context?

This is at once the most interesting of your questions, and the most difficult to answer. The changes to widely accepted views of Vermeer, necessitated by a recognition of his extensive use of the camera, are confined I believe just to certain aspects of his work, while others remain quite unaffected. The ostensible subject matter of Vermeer's pictures from 1657 onwards does not depart very greatly from the 17th century Dutch genre tradition in which he worked. We read the pictures as variants of standard themes - scholars in their studies, the artist in his studio, women reading love letters, 'merry companies' - although with Vermeer there is often a tranquillity and melancholy, where say in Jan Steen there would be rumbustiousness and chaos. Vermeer is never obvious, and it can be difficult to gauge the exact psychological relationships between his sitters, as in pictures like 'The Music Lesson' or 'The Concert'. His arrangements of figures, and the activities in which they engage, are nevertheless not so different from equivalent works by his contemporaries.  So far as allegory and symbolism go, Vermeer again conforms to accepted meanings, albeit with a large element on occasion of ambiguity and elusiveness. The 2001 New York/ London exhibition 'Vermeer and the Delft School' sought to situate Vermeer within his local artistic milieu in exactly these kinds of terms.  

For many visitors to that exhibition however, the curators' central premise was undermined by the fact that, set alongside Pieter de Hooch, the Delft architectural painters, even Fabritius, Vermeer did not seem to be a member of any 'school'. On the contrary, his pictures stood out by their strangeness and their radically different visual 'feel'.  This was not just because of Vermeer's superior powers as a painter, I would suggest. It was because for Vermeer his later pictures all have a second kind of subject matter: that is, the effects of light, shadow, colour and tone. All realist painters seek to capture these luminous attributes of appearances, of course; but for Vermeer they become subjects in themselves, as much as they are means to the representation of human faces or satin jackets or carpet-covered tables. It is in the response to these optical phenomena that the wider scientific and cultural significance of Vermeer's camera technique lies.  

In my book I quoted Lawrence Gowing who says that "Vermeer is alone in putting [the camera obscura] to the service of style rather than the accumulation of facts." For Vermeer, that is to say, the camera was not merely a tool for achieving correct perspective or for refining composition. It was both these things, certainly; but much more important, it was the key to a whole new way of seeing. The camera was, for Vermeer, an instrument through which to gain a new vision of the world at the scale of the everyday, just as the telescope and the microscope were instruments for gaining new views of the worlds of the very large and the very small.  

Vermeer was a supremely intelligent painter, perhaps one should say an intellectual painter; but at the same time he had I believe an extraordinary capacity for switching between this intellectualised, rational eye, and what one might call a perfectly 'idiotic' eye, with which he was able to see luminous patches of hue and tone, quite independent of the real-world objects from which they emanated.  As Gowing puts it, in this mode of seeing, "Vermeer seems almost not to care, or not even to know, what it is that he is painting. What do men call this wedge of light? A nose? A finger? What do we know of its shape? To Vermeer none of this matters, the conceptual world of names and knowledge is forgotten, nothing concerns him but what is visible, the tone, the wedge of light." Aldous Huxley in his book The Doors of Perception describes the world seen under the influence of mescalin as being like a painting by Vermeer: "Things without pretensions, satisfied to be merely themselves, sufficient in their suchness, not acting a part, not trying, insanely, to go it alone..." The drug, as one might say, switches off the higher-level powers of conceptualisation, leaving the eye to see just light and colour - as the painter was able to do without the benefit of artificial stimulants.  

I was recently much struck by a passage from the 18th century Scottish philosopher Thomas Reid, who speaks of the general habit of the human mind to move extremely rapidly, from the reception of stimuli in the eye, to mental interpretations of those stimuli in terms of known objects in the world. "The mind has acquired a confirmed and inveterate habit of inattention to [the luminous stimuli]; for they no sooner appear than quick as lightning the thing signified succeeds, and engrosses all our regard..." The only profession in life in which it is necessary, by training the eye and mind, to break this process apart - to separate seeing from recognising - Reid says, is painting. "The painter hath occasion for an abstraction, with regard to visible objects... and this is indeed the most difficult part of his art. For it is evident, that if he could fix in his imagination the visible appearance of objects, without confounding it with the things signified by that appearance, it would be as easy for him to paint from the life, and to give every figure its proper shading and relief, and its perspective proportions, as it is to paint from a copy."  

Once again, all naturalistic painters must have this skill, to an extent; and it is an important part of their training to cultivate it. But Vermeer was able I believe to go much further than his contemporaries in this direction, through his use of the camera obscura. The camera, after all, serves to collapse the appearances of the three-dimensional world onto the two-dimensional surface of the camera screen. It allows the painter to inspect, directly, an 'optical array' of luminous points and patches, resembling the array of stimuli which fall on his retina. The fact that the camera image may be softened and simplified, because of slight loss of focus in the lens or the passage of light through a ground glass screen, can further help him to make this mental separation between image and reality. Vermeer's genius was to recognise, if only perhaps intuitively, that if he could match in pigment the relative tonal values of these luminous patches, then the observer's intelligent eye would do the rest. He did not have to be particularly precise in outline, so long as he was 'uncannily' true to tone.  

Readers of the Essential Vermeer site will need no reminding that our knowledge of Vermeer's public life and professional training is minimal, and that we have no information whatsoever - other than what can be gleaned from the paintings - about his mental life or the extent of his learning. We are therefore reduced to speculation here. It is possible all the same, I think, to see in Vermeer's use of the camera obscura some intellectual connections to, or parallels with, those other 17th century scientific discoveries which were made possible by lenses. Knowledge and mental training were required, in order to interpret the phenomena which were made visible by the microscope and telescope, and which had never previously been accessible to the human eye.  Distinguished visitors who came to look through Galileo's telescope complained that they could make out nothing but areas of dark and light. The same happened to Leeuwenhoek's guests, when they tried out his microscopes. The problem was that these observers possessed no mental categories, no prior expectations, by which they could organise such novel perceptual data. It needed Galileo's scientific insight - aided by his training and skill as a draughtsman - to read the darks and lights of the telescopic image as the craters and mountains of the moon.  In the same way, Christiaan Huygens was able to interpret Saturn's strangely fluctuating appearances as a series of perspective views, from different angles, of the planet and its rings.

For Vermeer the situation was the other way round. He could rely on the viewers of his paintings to have all the mental constructs and expectations needed for recognising the familiar objects in Dutch domestic interiors. His achievement was to recover and duplicate in paint, with the aid of the camera obscura, the initial optical stimuli by which those expectations could be triggered. This theme, of Vermeer's place in the wider world of 17th century scientific thought and experiment, is developed in Robert Huerta's fascinating new book Giants of Delft, to be published in June 2003 by Bucknell University Press.

In your opinion, has your educational background consciously or unconsciously influenced the negative reaction to your theories on the part of those museum curators and Vermeer scholars with art history backgrounds?

The great majority of the reviews of Vermeer’s Camera, many of them by professional art historians, have been enthusiastic. You talk about a ‘negative reaction’. The book had just two critical notices, from Jørgen Wadum and Walter Liedtke; and I have also experienced scepticism or worse in some seminars and conferences. From other directions where one might perhaps have expected reactions there has been silence. I am not sure however whether it is my background as an architect, as such, that has provoked hostility or indifference. After all, there have been other contributions to Vermeer studies by ‘outsiders’, notably John Michael Montias, which have been welcomed by the art-historical profession.

I would think that any resentment might have more to do with my presenting a serious challenge to accepted views and entrenched opinions. What is notable is that those art historians (and others) who have been most receptive are those with a real understanding of perspective geometry – which is I my own field of expertise – and perhaps also with a foot in the history of science. By contrast Wadum and Liedtke simply fail to address, and show little evidence of really comprehending, the central geometrical argument of Vermeer’s Camera. They concentrate instead on trying to find fault in minor and peripheral points. Some of their criticisms are simply mistaken, or are answered in the book.

What are your feelings about the David Hockney-Charles Falco hypothesis of undiscovered and widespread use of optical instruments in European painting from the fifteenth-century on?

This is a hard question to answer briefly, not least because the 'David Hockney - Charles Falco thesis' is very wide-ranging and ambitious, but at the same time Hockney and Falco have not always made clear precisely what they are claiming in relation to any given artist, school or period.  I am of course entirely convinced that Vermeer used an optical method (although I am not sure whether many of his Dutch contemporaries did - the most likely candidates being van Hoogstraten and Dou). There are some very provocative questions raised by Hockney and Falco in relation to Caravaggio, who certainly had access to optical expertise and equipment. These deserve further investigation. There can also be no doubt that in the 18th and 19th centuries the use of the camera obscura by artists was widespread. And Hockney's suspicions that Ingres was an early user of the camera lucida appear now to be well-founded.

What is more problematic is Hockney and Falco's proposal that from the early 15th up until the late 16th century, numerous artists both in Italy and in Northern Europe studied and traced images projected by concave mirrors. (Hockney calls them 'mirror-lenses'.) There is very little supporting documentary evidence, and Hockney's contention that this was 'secret knowledge' which artists kept to themselves is, like all conspiracy theories, intrinsically untestable.  Leonardo da Vinci knew all about concave mirrors, but nowhere mentions their use for picture-making - something which he, of all people, would surely have known about, if it was general practice among painters. And as J V Field has pointed out, if artists did use 'mirror-lenses', it is also surprising that neither Galileo nor Kepler - both of whom were knowledgeable about and skilled in drawing - make any mention in print.  

On the other hand, it seems to me that there is a degree of plausibility in what one might call a 'weak version' of the Hockney-Falco thesis as applied to Renaissance painting. This would draw looser connections than they do between the emergence of naturalism, and the greatly-improved quality of Italian-manufactured glass mirrors - both curved and flat - at this period. A very suggestive argument along these lines is offered by Alan Macfarlane and Gerry Martin in their recent book The Glass Bathyscaphe: How Glass Changed the World (Profile Books 2002). It seems very clear to them (and to me) that Brunelleschi was led to the discovery of linear perspective through consideration and measurement of the images formed in plane mirrors. Filarete said as much at the time. Once painters became interested in naturalistic appearances, then it is only to be expected that mirror-images of faces would come to influence portraiture; and that high-quality glassware generally would provoke interest among painters in the phenomena of refraction, reflection and specular highlights.

One does not have to imagine, in positing such influences, that all artists used optical devices - even as simple ones as plane mirrors - in their own practice. Martin Kemp writing in Hockney's Secret Knowledge sets out a 'taxonomy' of uses of optical instruments by artists, which is helpful in this connection. Those uses could range from 'literal copying of contours', through 'imitation of line, colour and tone' as seen in optical images, to 'aping' of the visual effects of lens-based devices, to 'faking it' - 'that is to say knowing what a camera image can deliver, and using its 'lessons' to paint/ draw freehand'. Beyond this, one might suppose that, once a great artist (like Vermeer) had learned from optics and made use of his knowledge in painting, then other followers could learn to achieve comparable effects by studying his pictures, without necessarily having to use the same apparatus themselves.

Do you think that the David Hockney-Charles Falco study has hindered or has been helpful in the acceptance of your own arguments in regards to Vermeer's use of the camera obscura?

In one very practical sense, the Hockney-Falco thesis has greatly helped acceptance of my ideas. Secret Knowledge was published shortly after Vermeer's Camera and attracted enormous attention, which spilled over to my book. On the other hand, there are weak points in Hockney and Falco's general position, as I have indicated; and these have, in some quarters, been used as the excuse for wholesale dismissals of all arguments for the use of optical devices by painters, in all places and periods. Let us hope that in the future we can move beyond such an unhelpful polarisation between supporting and opposing camps in this debate, and consider individual cases, and much more focussed arguments, on their respective merits.

survey software

need + information?

poll initiated: march 12, 2014